Vocabulary Patterns in Free-for-all Collaborative Indexing Systems

Transcription

1 Vocabulary Patterns in Free-for-all Collaborative Indexing Systems Wolfgang Maass, Tobias Kowatsch, and Timo Münster Hochschule Furtwangen University (HFU) Robert-Gerwig-Platz 1, D-781 Furtwangen, Germany Abstract. In collaborative indexing systems users generate a big amount of metadata by labelling web-based content. These labels are known as tags and form a shared vocabulary. In order to understand the characteristics of that vocabulary, we study structural patterns of these tags by implying the theory of self-organizing systems. Therefore, we utilize the graph theoretic notion to model the network of tags and their valued connections, which represent frequency rates of co-occurring tags. Empirical data is provided by the free-for-all collaborative indexing systems Delicious, Connotea and CiteULike. First, we measure the frequency distribution of co-occurring tags. Secondly, we correlate these tags towards their rank over time. Results indicate a strong relationship among a few tags as well as a notable persistence of these tags over time. Therefore, we make the educated guess that the observed collaborative indexing systems are self-organizing systems towards a shared vocabulary building. Implications on the results are the presence of semantic domains based on high frequency rates of co-occurring tags, which reflect topics of interest among the user community. When observing those semantic domains over time, that information can be used to provide a historical or trend-setting development of the community s interests, thus enhancing collaborative indexing systems in general as well as providing a new tool to develop community-based products and services at the same time. Key words: Metadata, tagging, shared vocabulary, online community, collaborative software, self-organizing system 1 Introduction Cooperative, distributed labelling of content in the worldwide web is called collaborative indexing or social tagging. Within a collaborative indexing system users annotate different contents e.g.: events 1, video clips, music 3, pictures 4,

2 articles and references 5, weblogs 6 or websites 7. These collaborative indexing systems facilitate mass categorization establishing so-called folksonomies, which is a bottom up categorization made by a large user base. A collaborative indexing system has basically two features. First, it is used for future retrieval of self-indexed content. Secondly, it provides recommendations, which are based upon the co-occurrence of highly used tags within all annotations, whereas we call one single process of annotation an indexing task. 8 The recommendations are shown to the user by committing a tag query. For instance: content tagged with html will be frequently tagged with css as well. The data collected within an indexing task contains the name of the user, an url linking to the content, one or more tags and time-stamp information. Therefore, the data within a collaborative indexing system is basically a network of users, tags and content in a given period of time. All tags together represent the shared vocabulary of the user community. In this paper we study the structural patterns of that vocabulary, thus focusing only on the partial network of tags. Analyzing this partial network requires some constructs of the graph theory. We assume the shared vocabulary to be a self-organizing system by means of the systems theory [1]. Hence, stable patterns as well as specific correlations are determined throughout the vocabulary. In addition, implications on these patterns are presented. To support the requirements of self-organizing systems by reducing external restrictions and forces we choose the free-for-all collaborative indexing systems Delicious, Connotea and CiteULike for empirical data extraction, where any user can index any content element. Thus, indexing rights are not restricted as identified by Marlow et al. []. This paper starts with related work covering collaborative indexing systems and the systems theory. Then, we hypothesize two assumptions regarding stable patterns within the vocabulary. Afterwards, we build up a model based on the graph theoretic notion, clarify the methodic approach and present the empirical data used to prove the assumptions. Subsequently, we present and discuss the results of our analysis and draw implications on them. Finally, we give an outlook on further research. Related Work A general review on collaborative indexing systems is given by Voss [3]. Mathes [4] discusses the organization of information via tags and points out that user generated metadata is of an uncontrolled nature and fundamentally chaotic compared to a controlled vocabulary. But he also mentions that collaborative index There may also exist other recommender implementations, but we focus on the cooccurrence of highly used tags because this information is freely accessible on the web.

3 ing systems are highly responsive to the users needs and their vocabulary by involving them into the process of organization. Vander Wal [5] distinguishes between broad and narrow folksonomies depending on the amount of users, who tag one specific content element. He also defines the difference between pure tagging and folksonomy tagging. Voss [6] discovers power law distributions of tag frequency rates in Delicious and Wikipedia supporting the presence of self-organizing systems. Hotho et al. [7] and Quintarelli [8] find power law distributions according to collaborative indexing systems, too. Lund et al. [9] measure a power law distribution of user shared tags within Connotea. Results of Golder and Huberman [1] show regularities of dynamic structures within Delicious. Moreover, they introduce a classification on the semantics of tags as well as Zhichen et al. [11]. Wu et al. [1] distinguish the potential of collaborative indexing systems as a technological infrastructure for acquiring social knowledge. Millen et al. [13] study the deployment of a collaborative indexing system within a company and highlight the remarkable acceptance rate of the users as well as its personal and organizational usefulness. In addition, Damianos et al. [14], Farrell and Lau [15] as well as John and Seligmann [16] also examine the potential of collaborative indexing systems for the enterprise covering people s expertise, social networks and the integration of those systems in existing collaborative applications. An early classification of collaborative indexing systems is done by Hammond et al. [17] confronting scholarly and general resources with links and web pages. In a more detailed classification Marlow et al. [] distinguish the design of a system and present several user incentives. Heymann and Garcia-Molina [18] develop an algorithm, which generates a hierarchical taxonomy of a tag network. For the same purpose Mika [19] uses social network analysis on the network of users, tags and content. Hotho et al. [7] develop a search algorithm for folksonomies to find communities of interest within collaborative indexing systems. Cattuto et al. [] design a stochastic model for the analysis of indexing tasks over time consisting of tags and users. Dubinko et al. [1] visualize tags over time with data from Flickr, whereas Zhichen et al. [11] propose an algorithm for tag suggestions to support the user within an indexing task. An overview of self-organizing systems is given by Heylighen [1]. 3 Motivation As mentioned above, this paper deals with the partial network of tags. The concept of tags is central in collaborative indexing systems. The same tags used by different users to annotate similar content show a common understanding of the users. The set of all tags utilized by the user community represents the shared vocabulary. Users and content elements are linked to each other through tags, which are also directly connected when they are used together within one indexing task. Figures 1 and are representing such an indexing task as well as the resulting network of the tags sports, worldcup and soccer. Due to the current work, the value of those tag connections is an essential dimension, which

4 is based on the frequency rate of tags co-occurring within all indexing tasks. A prerequisite for a measurement of this frequency is the bag-model for aggregation of tags, in which multiple tags can be assigned to one resource by multiple users as discussed by Marlow et al. []. worldcup url description FIFAworldcup.com The Official Site of FIFA World Cup tags sports worldcup soccer Fig. 1. Graphical input mask for an indexing task soccer sports Fig.. Resulting network of the indexing task in Fig. 1 Prior work on stable patterns suggests that collaborative indexing systems are self-organizing systems [1,, 6, 8, 9]. The vocabulary - consisting of tags and generated within all indexing tasks by all users - is a part of this system, which organizes its structure by itself, without a centralized control mechanism. The users of a collaborative indexing system generate this vocabulary in a decentralized approach, not even aware of it. On its own this system evolves over time into a more stable state. Contrary to the aforementioned work, we explore patterns emerging out of cooccurring tags. Therefore, we want to know if the power law distribution, which is common in broad folksonomies [7, 9, 8], is also applicable to the structure of co-occurring tags. This would represent a community s vocabulary, which consists of a few tags co-occurring with high frequency rates and many tags cooccurring with low frequency rates. Such a pattern - we call it tag economics - would indicate a strong consensus on a particular subpart of the community s vocabulary, from which particular interests of the users can be identified. Due to these considerations, we hypothesize the relation of co-occurring tags as follows: H1 Let T i be a tag and T j i all tags co-occurring with T i. Then the ranked frequency distribution of all valued connections from T i to T j i follows a power law curve. Additionally, we focus on the frequency dynamics of tags over time depending on their position in the aforementioned frequency distribution. We assume that tags co-occurring with high frequency rates (higher position on the power law curve) are more stable over time than tags co-occurring with low frequency rates. This would represent persistence of the community s interests or, when tags with high frequency rates change to a low position, one can suggest a shift of the community s common understanding. Therefore, the current work has the second objective to examine the relationship of the frequency rates of cooccurring tags over time. We hypothesize this relationship as follows: H The higher the frequency rates of the tags T j i, the more stable are they over time.

5 4 Model Let an indexing task be a quadruple comprised of < user, url, timestamp, tag >. One user enters an url with none, one or more tags into a collaborative indexing system at a certain time. Only two entities are important according to our hypotheses, namely timestamp and tag. Therefore, The community s vocabulary is modelled as an undirected, valued and finite graph V within a given period of time δ. This period of time is essential, because the frequency of time based indexing tasks is subject to fluctuations, which occur in the course of a day, a week or month. Furthermore, δ can be used to affect directly the size of the vocabulary V to ease the analysis. The vocabulary V consists of a set of nodes (here tags) and a set of valued links, which represent the frequency values of co-occurring tags. Hence, we refer to this vocabulary as the network of tags, too. The links are undirected since each tag i, which co-occurs with a tag j, also means that the tag j co-occurs with the tag i, respectively. To better handle these frequency values, the vocabulary can be described by a symmetric frequency matrix F, such that the value on the ith row and jth column represents the frequency rate of the co-occurring tags i and j over all indexing tasks within δ, denoted as f(i, j). Self references are excluded since we focus only on co-occurring tags. Thus, the diagonal values f(i, j) with i = j are always zero. Figure 3 exemplifies an undirected, valued graph of the vocabulary V, whereas Fig. 4 shows the corresponding frequency matrix F. Based upon this graph theoretic notion and the corresponding frequency matrix, we are able to illustrate and compute the frequency distribution of co-occurring tags Fig. 3. Undirected, valued graph of the vocabulary V including 5 tags i Fig. 4. Corresponding frequency matrix F of the vocabulary V 6 3 j 4.1 Method A frequency matrix F (δ 1 ) is built within a given period of time. Afterwards, the frequency values f(i, j) for each tag T i are summed up. Consecutively, those

6 cumulative frequency rates are ranked by size and confined by a limit L. This approach eliminates tags T i with low cumulative frequency values of co-occurring tags T j i, because they cannot contribute any meaningful values for co-occurring tags and are therefore not relevant for further calculations. Then, N tags T i with maximum N max, medium N med and minimum N min cumulative frequency rates are identified. Afterwards, the frequency distribution of all tags T j i co-occurring with each tag T i is calculated from this selection and subsequently ranked by size. Finally, the values of these frequency distributions are normalized and utilized to conduct a curve estimation regression statistic based on the power model, whose equation is f(r) = β r β1 with f(r) estimating the frequency rate depending on the frequency rank r of a tag T j i. The results of the regression statistics are used to prove hypothesis 1. The aforementioned N tags T i are also used to prove the second hypothesis. Hence, N tags T i with maximum N max, medium N med, and minimum N min cumulative frequency rates are identified. Afterwards, the frequency rates for each pair of co-occurring tags are written down in a time series each lasting δ over I iterations. To measure the stability between co-occurring tags T i and T j i, the difference D from the mean frequency of each tag co-occurrence in N max, N med, and N min is calculated over all I iterations, so they can be compared afterwards. 4. Empirical Data The empirical data for the analysis was extracted from the collaborative indexing systems Delicious, Connotea and CiteULike. This information is freely accessible. Indexing rights are based on a free-for-all principle [], thus supporting the requirements of self-organizing systems by reducing external restrictions and forces. The content is respectively of textual nature. The selected collaborative indexing systems differ in the community s size and the quantity of indexing tasks, the amount of tags, as well as the period of time in which the data was gathered. Furthermore, all indexing systems use a bag-model to aggregate tags, which is essential for our approach as mentioned in Sect. 3. Table 1 provides detailed information about the empirical data. Table 1. Empirical data Indexing system Delicious Connotea CiteULike Period of measurement 9/1/6 1/1/6 1/1/6 1/1/6 9/17/6 1/1/6 Indexing tasks (It) It incl. at least tags (6%) (61%) 43 (5%) Tags incl. doublets Distinct tags (11%) (17%) (37%) Distinct users Users with at least It (58%) 7 (69%) 48 (64%)

7 5 Results 5.1 Hypothesis 1: Power Law Distribution of Co-occurring Tags The ranked frequency distribution f(r) of T j i tags co-occurring with a tag T i is illustrated in Fig. 5. Thus, a power law distribution is clearly apparent in the shared vocabulary, as to be expected from a broad folksonomy like Delicious. There are many tags T j i with low frequency rates of co-occurring tags and few with very high frequency rates. This result is proved by the cumulative discrete co-occurrence distribution in Fig. 6, which illustrates the discrete frequency distribution. There is a remarkable gap between tags, which co-occur with only one single tag, and tags co-occurring with multiple other tags. Towards the high co-occurrence rates the curve decreases rapidly as the logarithmic scale demonstrates. Frequency rates of co-occurring tags above 1 lead to absolute frequency rates less then ten. Similar results are provided by the collaborative indexing systems Connotea and CiteULike. The visual observations in Fig. 5 and 6 can be confirmed statistically. Therefore, Table shows the median of squared reliability ( R ), the median degree of freedom ( F ) as well as the exponent β 1 according to the power law curve estimation algorithm 9 over all corresponding tags within N max, N med, and N min. Compared with other curve estimation algorithms, the power model performed best by far. In particular, data from Delicious with 7 co-occurring tags T j i as a median degree of freedom shows a remarkable reliability of.96 by only.1 standard deviation for N max. Values with less reliability values lie nearby.8, which is still acceptable although standard deviation values show higher dynamics. Additionally, a decrease of the exponent β 1 can be observed related to the degree of freedom by considering the data of Delicious and Connotea. This can be referred to a smoother power law curve, when less tags co-occur with a tag T i. For this reason, the relative low degree of freedom according to CiteULike can be neglected to identify the aforementioned effect. Due to these facts, the first hypothesis is supported by the empirical data. It is quite evident that a power law curve of co-occurring tags is obvious for tags T i with high frequency rates, whereas the co-occurrences of middle and low ranked tags T i show more dynamics. 5. Hypothesis : Relation between Rank and Persistence of Tags over Time Figure 7 shows the dynamics over time of co-occurring tags against their deviations from the mean frequency. As illustrated, co-occurring tags with high frequency rates - values from Nmax T1 5 - are more stable and have a lower scatter respectively than co-occurring tags from N med or N min. Figure 8 shows the 9 Statistical software used for curve estimation: SPSS, Version 15..1, SPSS Inc. Chicago, USA

8 f(r) r Fig. 5. Extract from the ranked frequency rates f(r) of T j i tags co-occurring with 3 tags T i (different symbols) based on Delicious, δ 1: 9/1/6 9/19/6 1 1 frequency discrete co-occurrence distribution Fig. 6. Discrete co-occurrence distribution of tags T i and T j i from Fig. 5 and the corresponding frequency values

9 Table. Curve estimation regression statistics based on the power law f(r) = β r β 1, R : median of squared reliability, F : median degree of freedom, standard deviations are provided in brackets Indexing system Delicious Connotea CiteULike N / L 3 / 5 3 / 5 / 1 δ 1 9/1/6 9/19/6 1/1/6 9/5/6 9/15/6 9/5/6 R / F / β 1 for Nmax.96 (.1) / 7 / -.96 (.8).93 (.1) / 638 / -.8 (.8).81 (.1) / 58 / -.46 (.18) for N med.8 (.8) / 15 / -.51 (.1).86 (.6) / 88 / -.58 (.9).79 (.1) / 3 / -.47 (.3) for N min.78 (.11) / 73 / -.4 (.1).84 (.11) / 55 / -.53 (.8).8 (.) / / -.65 (.4) relative frequencies of two co-occurring tags from N max and N med in contrast to each other over 3 iterations with δ = 1 day. This figure illustrates that the interval of N med shows higher variations than the interval of N max. The basic data from all examined collaborative indexing systems with the average deviation of the mean frequency values D over N max, N med, and N min is shown in Table 3. As a result, δ and the number of indexing tasks within δ are affecting the dynamics. Those in Connotea and CiteULike are much lower than the dynamics in Delicious. This often causes co-occurring tags to appear only in a small degree over all iterations and therefore, they stabilize only on a low level with low frequency rates. Thus, a small deviation from the average frequency rate over all iterations is not always indicating a high position of co-occurring tags on the power law curve. There are also situations, where tags stabilize on low frequency rates. The stability alone is therefore not a sufficient criterion for the occurrence of high frequencies. In fact, the absolute frequency rates must be observed for a positioning on the power law curve besides the deviation. As a result, correlations between high frequency rates and their persistence over time can be concluded, but not vice versa. A change of that persistence is therefore only significant for a shared vocabulary, if the deviation of the average value appears on high frequency rates. Nevertheless, the second hypothesis is also supported by the figures of Table 3, although with less explanatory power compared to the findings of hypothesis 1. Table 3. Deviation over the average frequency Indexing system Delicious Connotea CiteULike N / L 5 / 3 3 / 1 15 / 5 δ 1 day, 1 month, 1 day, 9/1/6 9/3/6 1/1/6 8/31/6 9/16/6 9/3/6 D(Nmax) 14.1%.% 15.1% D(N med ) 19.4% 4.% 16.1% D(N min ) 19.8% 7.% 16.7% 5.3 Implications As shown in section 5.1, we observe a frequency distribution of co-occurring tags which follows a power law curve. The examined collaborative indexing systems

10 .3 Deviation over the average frequency in % T i Fig. 7. Deviation over the average frequency in % of 75 tags T i and all related tags T j i for N T 1 5 max, N T 51 5 med and N T min based on Delicious, between 9/1/6 9/3/6 relative frequency in % day Fig. 8. Relative frequency comparison in % of 1 high (dashed) and low positioned tag T i from Fig. 5 with δ = 1 day and 3 iterations based on Delicious, between 9/1/6 9/3/6

11 support a distributed approach without central control mechanisms, favoring self-organization. The observed distribution of tags means that the users have a strong consensus at least on a particular subpart of the shared vocabulary, since co-occurring tags with high frequency rates build a semantic domain. This shows some sort of tag economics within a collaborative indexing system. A further aspect indicating a self-organizing system is the resilience of the system. Accidental errors, e.g., typos or willful sabotages of the system by users have negligible effects, because single users cannot tip the scales of a power law curve. In addition, the construct of indexing support through pre-defined tags, which is suggested to consolidate the tag usage [11, ], would additionally support these findings by diminishing the limits of the uncontrolled vocabulary such as polysemy, synonymy/uniformity and basic level variation problems [1, ]. Hence, we suggest higher frequency rates within the top ranked tags as well as lower rates within low ranked ones as fundamental impacts of the indexing support construct. Another feature of a self-organizing system is the adaptation of environmental changes. In terms of a collaborative indexing system, these changes can be referred to as a shift of the community s interest, which is likewise reflected in a structural change of the vocabulary. Hence, semantic domains based upon cooccurring tags with high frequency rates may change. For instance, if the position of a tag T j i alters over time by means of an increase or decrease of the frequency rates according to T i, then this progress suggests a structural change within the vocabulary and vice versa. The higher the position of this tag on the power law curve, the more significant is the structural change of the vocabulary. When this dynamic information is monitored one can observe a historical or trend-setting development of the vocabulary based upon the time-stamp of selected indexing tasks. Those trend curves of the vocabulary suggest changes within the community s interest and are useful for the particular user when searching for content elements, users or tags in the time domain. 6 Conclusion and Future Work In this paper we studied structural patterns of user generated vocabularies within the free-for-all collaborative indexing systems Delicious, Connotea and CiteU- Like. The theory of self-organizing systems was implied to hypothesize patterns within those vocabularies. We built up a model based on the graph theoretic notion consisting of tags and their valued connections. This was required to calculate the frequency distribution of tags that co-occur with others, as well as to correlate those tags towards their frequency rate over time. Results indicate that only a few co-occurring tags exist with high and many with low frequency rates, thus following a power law curve. In addition to that, co-occurring tags with high frequency rates proved to be more stable over time than those with low rates. The results were also depending on the quantity of indexing tasks. For instance, the measured values of CiteULike yielded less explanatory power than the values of Delicious. Implications are drawn through

12 the presence of semantic domains, which are based on co-occurring tags with high frequency rates and the shift of common interests among the user community, if those high rates are fundamentally changing over time. The resulting information can be used to provide a historical or trend-setting development of the vocabulary and would not only be useful for the particular user but would also support enterprises to develop products and services, which may depend on or at least involve the interests and trends of online communities. Due to the current work, the development of algorithms for trend information and historical time series based on the frequency distribution of co-occurring tags is an interesting area for further research. A common understanding of the user community is expressed through the tag network comprised of valued links with high frequency rates. In addition, semantic domains of more than two co-occurring tags can also be identified with techniques of the social network analysis such as centrality measurements or clustering. This network alters dynamically in a self-organizing way over time suggesting new topics or events of social, academic, technical or economic nature. Defining triggers on the observed power law curve to identify those variances requires further clarification, but would be very useful by supporting users, enterprises or public organisations in upcoming decisions. Moreover, there is still a challenge in collaborative indexing systems featuring low indexing rates within a given period of time. This applies especially for those systems deployed in companies. Therefore, it is essential to find techniques, which permit major vocabulary coherence in such minimal systems and boost the significance of the common understanding. References 1. Heylighen, F.: The science of self-organization and adaptivity. In: The Encyclopedia of Life Support Systems, Oxford, UK, Eolss Publishers (1999). Marlow, C., Naaman, M., Boyd, D., Davis, M.: Ht6, tagging paper, taxonomy, flickr, academic article, to read. In: HYPERTEXT 6: Proceedings of the seventeenth conference on Hypertext and Hypermedia, New York, ACM Press (6) Voss, J.: Tagging, folksonomy & co - renaissance of manual indexing? ArXiv Computer Science e-prints (January 7) 4. Mathes, A.: Folksonomies - cooperative classification and communication through shared metadata. Technical report, Graduate School of Library and Information Science, University of Illinois (December 4) 5. Vander Wal, T.: Explaining and showing broad and narrow folksonomies (February 5) 6. Voss, J.: Collaborative thesaurus tagging the wikipedia way. ArXiv Computer Science e-prints (April 6) 7. Hotho, A., Jäschke, R., Schmitz, C., Stumme, G.: Information retrieval in folksonomies: Search and ranking. In Sure, Y., Domingue, J., eds.: The Semantic Web: Research and Applications. Volume 411 of LNAI., Heidelberg, Springer (June 6)

13 8. Quintarelli, E.: Folksonomies: power to the people. Incontro ISKO Italia - UniMIB Meeting (June 5) 9. Lund, B., Hammond, T., Flack, M., Hannay, T.: Social bookmarking tools (ii) a case study - connotea. D-Lib Magazine 11(4) (April 5) 1. Golder, S.A., Huberman, B.A.: Usage patterns of collaborative tagging systems. Journal of Information Science 3() (April 6) Zhichen, X., Yun, F., Jianchang, M., Difu, S.: Towards the semantic web: Collaborative tag suggestions. In: Collaborative Web Tagging Workshop, WWW 6, 15th International World Wide Web Conference, Edinburgh, IW3C (May 6) 1. Wu, H., Zubair, M., Maly, K.: Harvesting social knowledge from folksonomies. In: HYPERTEXT 6: Proceedings of the seventeenth conference on Hypertext and Hypermedia, New York, ACM Press (6) Millen, D.R., Feinberg, J., Kerr, B.: Dogear: Social bookmarking in the enterprise. In: CHI 6: Proceedings of the SIGCHI conference on Human Factors in computing systems, New York, ACM Press (6) Damianos, L., Griffith, J., Cuomo, D.: Onomi: Social bookmarking on a corporate intranet. In: Collaborative Web Tagging Workshop, WWW 6, 15th International World Wide Web Conference, Edinburgh, IW3C (May 6) 15. Farrell, S., Lau, T.: Fringe contacts: People-tagging for the enterprise. In: Collaborative Web Tagging Workshop, WWW 6, 15th International World Wide Web Conference, Edinburgh, IW3C (May 6) 16. John, A., Seligmann, D.: Collaborative tagging and expertise in the enterprise. In: Collaborative Web Tagging Workshop, WWW 6, 15th International World Wide Web Conference, Edinburgh, IW3C (May 6) 17. Hammond, T., Hannay, T., Lund, B., Scott, J.: Social bookmarking tools (i): A general review. D-Lib Magazine 11(4) (April 5) 18. Heymann, P., Garcia-Molina, H.: Collaborative creation of communal hierarchical taxonomies in social tagging systems. Technical report, Computer Science Department, Stanford University (April 6) 19. Mika, P.: Ontologies are us: A unified model of social networks and semantics. In: 4th International Semantic Web Conference (ISWC 5). (5). Cattuto, C., Loreto, V., Pietronero, L.: Collaborative tagging and semiotic dynamics. ArXiv Computer Science e-prints (May 6) 1. Dubinko, M., Kumar, R., Magnani, J., Novak, J., Raghavan, P., Tomkins, A.: Visualizing tags over time. In: WWW 6: Proceedings of the 15th international conference on World Wide Web, New York, ACM Press (6) 193. Furnas, G.W., Landauer, T.K., Gomez, L.M., Dumais, S.T.: The vocabulary problem in human-system communication. Communications of the ACM 3(11) (1987)